Low frequency Triangle generator

[HarveyH42]

I'm looking for a very low frequency triangle, 2-3 seconds between peaks. I tried a 555 timer, but looks more like a trapazoid, a lot of full on time. This is for fading some LEDs, so it got a little confusing with opamps. I don't want the LEDs to off too long before the voltage gets high enough to light them up, so need to start around 1 volt 20mA, and rise to about 3 volts 20mA.

There will be 3 of these triangle generators, of slightly different frequency, so need to be able to adjust them. Figure a quad opamp would be great, single chip, few caps few resisters... Unfortunately, I know very little about opamps as frequency generators, the examples I've been looking at ar for much higher frequencies, and my simulator doesn't show much oscilation (got to buy a better program soon).

Anyway, I have a few LM324, and an assortment of othe LM types (singles and duals mostly). Also some 339s, but don't think comparators can be used the same way. Also would like to keep it to a single power source, 9-12v batteries.

Basically, need an example to work from for a low frequency triangle wave. Doesn't have to be an opamp circuit, just seems the most common. Would rather use common parts (what good is a junk box if you never use the parts).

 

[audioguru]

Look at Fading Red Eyes on this page:
**broken link removed**

 

[HarveyH42]

Thanks, will build one tomorrow with a LM324 (know I got some). I ran a quick simulation, works, but not as slow as I was looking for and a little unstable. Could be just the program, not the best. Tried changing a few values, with some improvement, so figure experimenting further on a breadboard would be a better indication. Good description with the circuit, but need to study it a little closer.

Thanks again.

 

[audioguru]

I ran a quick simulation, works, but not as slow as I was looking for and a little unstable.

Use two back-to-back 100uF , 220uF or 470uF capacitors for the ramps to be much slower. Add a 100uF supply bypass capacitor for improved stability.

 

[HarveyH42]

Before I get started on the breadboard, I need to remember what to do with the unused opamps. Going to use an LM324 quad, and only going to use two of the opamps to start with (keep it simple in the beginning). I remembered something about unused inputs, perhaps I'm confusing it a little with logic chips, but pretty sure it's something different. Flipped through a few books, but didn't find anything to spark a memory.

Anyway, going to do some searching and see if anything turns up...

 

[HarveyH42]

Wow, found it first try. Books are useless?

"Unused op-amps in dual and quad packs should be terminated by grounding the + input and connecting the - input to the output."

 

[HarveyH42]

Just finished the cross-fade version of the fading eyes, used 2 x 220 uF caps, 2-red and 2-blue LEDs . The frequency is fine, going to let it run for a while and check for temperature problems. Looks okay, colors don't blend as well as I had hoped, even using three difusers. The digital camera didn't catch the purples to well. I 'm waiting on some 70 degree super-flux LEDs, any day now. Maybe the wider angle will help.

Thank you Mr. Postal-Dude! Their here... Gotta go.

 

[HarveyH42]

Well that turned out disappointing, got everything except the Blue Piranha LEDs. Kind puts a snag in my projects. The Piranha base kind of sux for breadboarding, takes up about as much space as a trimmer, except it doesn't fit across the center. Oh well, guess I'll have to mount a few on perfboard and run wires.
Sent an Email to see what's up with the blues, but who knows. It's overseas, so basically screwed. Not a huge sum of money, just having to wait a week or two before I can get some matching blues. Hope I don't have to find a new supplier, this guys are real cheap, even shipping ($6.00 from China).

Made a different diffuser. Also used a piece of PVC pipe to cover the LEDs. Colors blend real well now. Now to tear it down, I build a two channel version, instead of cross-fading. Probably just mirror the existing, component tolerences should allow the two colors to drift out of sync....

 

[audioguru]

Years ago I made some fading LEDs with red, green and blue ones. They shine on the white ceiling and the colors combine slowly making all colors including white.

 

[HarveyH42]

Having problems with fader #2. I built the same circuit using the other two opamps on the 324. Doesn't fade, just full on. Went over everything several times, checked the data sheet to make sure the pins were right, everything should be cool. Same problem, so I swapped the output wires to the leds (two red two blue). Fader one still working fine on other color, # 2 still full on.
Instead of taking off from the voltage divider from fader #1, I put in another one (maybe not into sharing), no change.

Pulled out another LM324 and swapped them out, defective parts happen... No change.

Pull the parts from Fader #2, #1 still working good.
built #2 with seperate 324, and guess what... Same problem.

Don't know much about opamps, but figure I'm not doing something right with the second circuit. Anyway, enough of that for one day...

 

[HarveyH42]

Finally got it running. Stupid problem... I renumbered the pins for the other two opamps, but didn't notice that the inputs on one was inverted. I've got all three channels working now. As I had suspected, the component tolerances were enough to keep the channels out of sync, but not enough for one color to dominate. Cool part is that I'm not noticing any all off (black), or all on (white), though it should happen.

Now to draw up a PCboard design. Think I leave that for next weekend.

 

[audioguru]

Cool part is that I'm not noticing any all off (black)

The amplitude of each triangle wave must exactly match the forward voltage of the two LEDs in each circuit for them to turn off. Experiment with different values for the circuit's 100k resistor.

or all on (white)

Do you have a green LED in addition to the red and blue ones? Then the fading patterns should include white.

 

[HarveyH42]

Yes, Red, Blue, and Green. The three signals are at slightly different frequencies, so different brightness, different colors. Comes close to white, but usually has some tint.

 

[audioguru]

Comes close to white, but usually has some tint.

Then your red, green and blue colors aren't true. The red might have some orange tint, or the green might be a little yellow, etc.

 

[HarveyH42]

Actually, the signals were just out of phase. When I got home from work today, there was white and black periods for a while, then they seperated again. I should hook up the O-scope sometime and see how close the 3 frequency are to each other. Pity its only a single trace.

Red is 625 nM, 20 degrees, 14000 mcd, 150 ohm resistor.
Green is 525 nM, 20 degrees, 14000 mcd, 100 ohm resistor.
Blue is 470 nm, 20 degrees, 9000 mcd, 100 ohm resistor.